This episode of the Creative Discussion podcast features a recent debate between ITIF's Joseph V. Coniglio and Joel Thayer of the Digital Progress Institute on the question of whether the American Innovation and Choice Online Act (AICOA) would help or hurt innovation and consumer choice. The debate was moderated by veteran tech journalist Nancy Scola and is re-aired here with no substantive edits. Watch the live recording here.

Cases Mentioned

• United States v. Philadelphia National Bank, 374 U.S. 321 (1963)
• Spectrum Sports, Inc. v. McQuillan, 506 U.S. 447 (1993)
• United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001)
• Epic Games, Inc. v. Apple Inc., 67 F.4th 946 (9th Cir. 2023)
• United States v. Google LLC, No. 1:20-cv-03010 (D.D.C. Aug. 5, 2024)
• Federal Trade Commission v. Meta Platforms, Inc., No. 1:20-cv-03590 (D.D.C.))

Auto-Transcript

​Nancy Scola: Good morning, everybody. My name is Nancy Scola, and I'm a journalist based here in Washington, DC, who generally writes on technology, politics, and public policy. I'm thrilled to be here this morning to moderate this debate on the question: would the American Innovation and Choice Online Act help or hurt innovation and consumer choice?

If you're joining us today, there's a pretty good chance you know what the American Innovation and Choice Online Act but just in case, it's a bill that has existed in some form since 2021, and it would, in short, aim to stop the very biggest online platforms from unfairly giving preference to their own products and services over that of competitors.

The bill was reintroduced last month by senators Chuck Grassley and Amy Klobuchar, and if passed and becomes law, it's probably safe to say it'd remake how a pretty big chunk of the internet works. So, the question is: is that a good thing or a bad thing? And we have two great debaters here this morning to wrestle over that question.

First, arguing that the bill would hurt innovation and consumer choice, we have Joe Coniglio, senior counsel and director of antitrust innovation at the Information Technology and Innovation Foundation, our host this morning. And second, arguing that the bill would help innovation and consumer choice is Joel Thayer, the president of the Digital Progress Institute.

This is meant to be a free-flowing debate. I'll ask questions, but Joe and Joel are encouraged to politely, of course, engage with one another as we go along.

So, Joel, Joe, good morning. Thanks for being here. So, first let's get the important stuff out of the way. The acronym for this bill, A-I-C-O-A. You hear it pronounced 'ay-co-uh' and 'eye-co-uh' in Washington.

Which is it? Joel?

Joel Thayer: Oh, don't put that on me. It's... it depends on whether you have a Southern twang.

Nancy Scola: Okay. Joe?

Joseph V. Coniglio: 'Eye-co-uh' is fine with me.

Nancy Scola: AICOA, okay, great. So, substantively let's start at a high level and work our way down. So, Joe, Joel says... Your names are very similar, so you're going to have to listen for the nuance here.

Joel Thayer: I was going to point that out. I was like, "Man, this is going to be tough."

Nancy Scola: I know.

Joseph V. Coniglio: You can use Joseph if you want.

Nancy Scola: Joseph. Joseph, Joel says that AICOA is grounded in more than a century of proven US antitrust law, and as he puts it, the bedrock principle that a monopolist cannot leverage control of an essential marketplace to crush competition.

What is Joel missing?

Joseph V. Coniglio: Sure. Well, thanks, Nancy, for moderating, and Joel, thanks for being here. This is going to be a great discussion. I appreciate everybody who tuned into this event. So, let me just say I do think there are a number of ways that AICOA deviates from the Sherman Act, and in so doing you know, is actually going to pose harms to consumers and innovation.

So, I think I'll start by just outlining some of those differences with the Sherman Act, the existing antitrust law. So, the first is namely how AICOA applies its restrictions. AICOA deals with a number of practices that restrict a firm's unilateral behavior, as we call it. Self-preferencing, refusals to deal, stuff that's typically handled under Section 2 of the Sherman Act.

And Section 2 of the Sherman Act only applies restrictions on a firm's unilateral behavior if that firm has monopoly power or some dominant position with an intent to exclude competitors and competition. AICOA is very different than that. AICOA applies its obligations without any analysis of whether a firm has monopoly power or dominance.

Instead, it basically just counts the revenues of firms and counts the users for particular products. So, the problem with that is, you know, the Sherman Act approach is designed to say, "We're only going to restrict firms' unilateral behaviors if we think there's some potential to harm competition, because they're a monopoly, because they're dominant."

By not doing that, AICOA risks restricting behavior in markets where there really is no risk of harm to competition through a dominant position or monopoly power. And I'll give you an example. You know, Microsoft's Edge browser, for example. Microsoft would surely be captured by the revenue thresholds, and its browser is a very widely used product, likely to trigger, I think, the user thresholds.

But Microsoft doesn't have any market power in its browser, right? So, we'd actually be restricting Microsoft from competing in the browser market against Google and Apple simply because it satisfies these numbers rather than actually has dominance. So, that's one big difference between the Sherman Act and AICOA.

Another big difference, in my opinion, is the standards that AICOA applies to restrict these practices relative to the Sherman Act. So, with these sort of unilateral behaviors under the Sherman Act, there's a pretty clear process that you have to show if you're evaluating these claims. So, the first is that the government shows that these practices are going to result in any competitive harm, and then the defendant has the ability to do one of two things: It can deny the government's theory and say that's not going to happen for a variety of reasons, or it can generally present pro-competitive justifications for its behavior and say, you know, "This is why, even despite those harms, consumers, innovation, competition are going to be better off on average." AICOA doesn't do that.

So, for the restrictions that deal with self-preferencing, AICOA does require the government to show some harm, but it doesn't give the defendant the opportunity to deny that case and say, "No, I don't agree with that theory of harm," right? That's a big difference. For other practices, AICOA doesn't even require the government to show harm at all, right? Again, big difference from the Sherman Act. Instead, it puts the initial burden on the defendant to show why there's not going to be harm. And, you know, if you think about that, in the normal Sherman Act context, you're responding to a particular theory, right? But here, you have to prove the negative which is really, really hard, and certainly not how the Sherman Act works.

So, that's another big difference. Now a third difference I would highlight is on the pro-competitive justification point. So, as I said, the Sherman Act gives firms general ability to present pro-competitive justifications for these sorts of behaviors, right? AICOA doesn't do that. AICOA limits the justifications to very, very narrow cases, like privacy and security, and it also raises the standard that you need to prove even those more limited justifications relative to the Sherman Act, right?

So, for the Sherman Act, you have to show preponderance of the evidence, and they're not pretextual, basically, under Section 2. AICOA requires a higher clear and convincing evidentiary standard, and it also requires to show things like that the restraint is narrowly tailored, it's the least restrictive alternative, right? Basically what we'd call strict scrutiny in a constitutional analysis. So, it's a very, very high bar and very different from the Sherman Act, and it's going to end up chilling a lot of pro-competitive behavior. The last thing I'll point out is, a lot of the behavior that AICOA goes after is what we call leveraging, right?

So, a firm that has a position in one market and leveraging that to benefit another market. That's actually not something you can condemn through Section 2 of the Sherman Act, right? We have Spectrum Sports and others. A leveraging offense is only in violating Section 2 if you're getting a monopoly in that second market, right? So, you have a monopoly in one market, and you're using it to monopolize a second market. Just getting an advantage in that second market doesn't violate the Sherman Act. So, I guess in all of those ways, I would say that AICOA is very, very different from the way we typically do antitrust law, and I think in ways it's actually going to harm rather than help consumers and innovation.

Nancy Scola: Thank you so much. So, Joel, the question to you is, what has Joseph gotten wrong there? And if you could focus a little bit on this, the question of not requiring that companies are engaging in monopolistic behavior, because to a layperson like me, that's the one that kind of jumps out as, okay, that does seem a little bit different from some of the existing antitrust law we have.

Joel Thayer: Sure. There's a lot to respond to, but I'll try to break this down into three basic arguments. One is I posit that the market actually is broken, and so this idea that somehow this one particular measure is the watershed moment that we've all been waiting for, where big tech just kind of, falters, all the services go away, I don't think it's really grounded in the true economics or the reality of how individual users are using any one of these platforms, and you can look at that through a myriad of different benchmarks. One, there are several different court cases right now, especially on the liability decision, where we found folks like Google, even Apple to a certain degree, and Meta to a lesser extent, violating these statutes that Joe very eloquently laid out for all of us.

And the idea was that, if you found them liable for the law, obviously there should be some sort of accountability on their end. It turns out that really isn't the way it worked out. I mean, you look at the Google remedies decision and you're like, did they even break the law at all? I mean, if we're just going to basically admit that they may have done things to unlawfully obtain the monopoly, but then we can't do anything to remedy that because of some ephemeral idea that this will impact the future of AI. And to me, that is emblematic of what the issue is with respect to antitrust law as applied to some of these individual companies or even, even this sector.

And the reason I say that is, I think the fact that we have this very general statute, and so we've heard a few, right? The Sherman Act and such are general authority statutes. They're broad. They're supposed to cover everything and every type of interaction. And so that, to me, from a first principle standpoint, I don't think is very helpful if you are trying to, challenge or at least alleviate some of the concerns that we are seeing in the market and some of the concerns that the courts themselves are seeing in the markets. The real thing that I think that needs to happen here is that, you know, maybe it is time for Congress to refresh the record a bit.

And it's also, when you look at AICOA, it's not really all that different than how we've treated other major sectors such as the banking sector where you have systematically important platforms in this case, but systematically important institutions in the banking context, where we have acknowledged that these are widely used platforms, and these are not things that are just these small companies and these small upstarts. I mean, Google is no longer the company that was started in a Stanford basement. This is a multi- billion, even trillion-dollar enterprise that operates a whole host of different services that are constantly getting vertically integrated and constantly eating up different aspects of the market.

Even markets that they weren't the first movers in or even the innovator in. And so that, I think, is part and parcel of the concern. So, the second point I'll point out is that I really do appreciate the idea of Congress stepping up and creating more clarity and a framework by which we can evaluate these particular harms. Instead of allowing the court to basically play, and I apologize to any judges in the room, to play economist or play technologist. I think what you're seeing here is that the fact that you have this general applied statute, you are allowing a lot of deference to courts who do things that are a little bit outside their remit, which is sit, basically be an economist by the bench or a technologist by the bench.

And I think you're starting to see that a little bit with Judge Boasberg's opinion in the Meta case where he's talking about market definition, and he says market definition is not defined by Congress, so we have to apply all these economic principles. And he looks at all these things and basically says, "Yes, even if I agree that Meta is a monopolist in the social media context, look at all these other things that they're getting into, and maybe that alleviates some of the concerns that you're seeing in this market."

Now, that to me is antithetical or anathema to the idea of what the laws are meant to protect. We are meant to have a structured view of what harms were committed in which market. And it seems like this general applicability statute and the hyper- vertically integrated aspects of these markets are allowing courts to extend beyond just reading the jurisprudence and extend beyond just evaluating the facts at issue. They are bringing in their own ideas of economics and their own ideas of where the technology is going, and you saw that even more prevalent in the Google remedies case. So, having a specific authority statute that cabins it and tells courts exactly what Congress is concerned with is very helpful.

I think even, to a certain extent, Joe would agree that there are issues in this market, we can resolve them. But I think we just disagree on the exact mechanisms. The second point is, I really prefer a specific authority statute over a general authority statute, because I just think that adds more mission creep from judges, and even adds more mission creep from agencies. So, in some ways, this cabins the agencies like the DOJ, the FTC, to basically articulate this specifically what we're after. Now, the third point, which I think directly addresses the things that Joe was describing, is that I think AICOA actually applies basic standards and actually clarifies a lot of procedural issues that we are seeing in the courts.

And I have a few that I think would, that make this point pretty well. one is that it clarifies the Philadelphia bank case, where in the Philadelphia bank case, we, that's where we got our 30% is, already, presumptively illegal in terms of a merger. In this case, it seems like what Congress is trying to do is appreciate that the 30% exists and that's how they've been operating, when it comes to merger review.

But it implies that, well, actually, it states now that 35%, which is 5% above what Philadelphia Bank suggests, and even above what the DOJ currently views as having market power, as the overall standard. So already that standard is raised in favor of the companies to a certain degree but at least it clarifies that type of standard so that way we're not messing around and wasting either the court's time or the public's time on over-lawyering not even at the merit stage, at the procedural stage, on what is this market and what defines this market.

The other issue is that it's not simply, "Oh, you did this, that, you're a big person and you're this company and you did something wrong". Every single one of these of these harms are linked to a very specific standard, and that is basically whatever materially harms competition. So, it's not just that you self-preferenced. You have to do it in a specific way, and Congress didn't stop there. They didn't just say materially. They didn't throw that phrase in there and didn't define it. They also defined what materially harms competition, and it looks a lot like what we've defined in traditional antitrust review, which is a lessening of competition, which is pretty much the standard in terms of how you evaluate competitive harms.

The last bit is that the added affirmative defenses, which also allows, the company to say "Hey, DOJ, you got it wrong." So, all the things that Joe was describing, which I think would be very absent these affirmative defenses, would be a very strong case as to why AICOA would be not the right vessel. I think the affirmative defenses actually gives a little bit more credence to this idea that the AICOA's not trying to regulate the watershed. They are not trying to stop things that are actually pro-consumer. They are saying that even if we have made the case, even though at the front end that, because again, at the front end, you have to be able to defend the fact that this action materially harmed competition.

Let's say you think you got it wrong and we're still going to move forward anyway. They can go to the court and say, "Okay, well, they say they made these arguments, but they completely ignored all of these other things that benefit consumers." That part of the statute allows that type of affirmative defense, which is right now implied in all of these analysis with respect to the Sherman Act. But it's not actually explicit, and so I think having those explicit affirmative defenses really does allow us to not just cabin the type of harms that the government can say that the firm did, but also allows the firm to adequately defend itself when the agency gets it wrong.

And I think I said three, and I'm sorry. I'll do one more.

Nancy Scola: Yeah.

Joel Thayer: The last bit that I think is really helpful in AICOA, and I think this is probably true across most of the antitrust bar, is that these cases take too doggone long. They last for five years at minimum, just to get to a merits decision, potentially. Maybe three years if you're on an expedited timeline. But what AICOA does is say, "With these particular harms, with these specific issues that we see as acute harms in the market, and that we have demonstrated as being materially harmful to competition, we are going to give you an expedited review so that way we don't run into the issues that we described earlier."

So, part of the criticism that I have with respect to Judge Boasberg and his Meta decision, and also Judge Mehta and his remedy decisions in Google, is that part of it was just that the entire market sort of shifted, where it allowed for ancillary or even tertiary issues that have nothing to do with the actual case, to get added in there, and basically caused a lot of disruption in just the executing of justice and executing of the enforcement.

So, I think ultimately, those factors actually weigh in favor of having AICOA being a pretty solid guardrail. And look, I think we have to say this at the front end. This is a discussion draft. This is not over. There's a lot of back and forth that can happen, and there seems like there's a lot of fine-tuning that we can do, and I think even the legislators are open to that.

But I think at the front end, to say this is not how we do things, I respectfully disagree. I think what this does is codify some of the standards and clarifies some of the harms that we actually find as harmful. And in some ways, it even cabins what the government can and can't assert in court, so that way we don't have these broad assertions being made and these retribution type of antitrust cases from the government.

So, I think this actually narrows it and cabins it. I think that's ultimately why I'm supportive of this measure.

Nancy Scola: Okay. So, Joe, I'm sure you have thoughts. I want to focus a little bit on the question we have: whether AICOA hurts or helps innovation and choice online. And Joel mentioned this idea of what concerns of Congress is the bill actually meant to address?

And Joe, I wanted to ask you, I'm sure you can bake in your response to Joel in your answer here, but on this self-preferencing question, this is the concern, you know, in a very simple way, of a lot of members of Congress, that you have these big online platforms where you search on Google for flight information, it returns Google Flights at the top of the search results.

Is there any self-preferencing that these platforms are engaging in that's so egregious that you think, "Okay, the law does need to step in here"?

Joseph V. Coniglio: So, look, the Sherman Act is already well-equipped to address instances of self-preferencing when a monopoly is using its market power to create another monopoly through technological tying or something like that.

That was what the Microsoft case was about all those years ago. So, there is law in the books to handle that. The problem with AICOA, of course, is that it, would go after self-preferencing even where there is no demonstration, that there will be a dominant firm or monopoly created, only if there's some de minimis harm to competition, which by the way, I would note the material harm to competition part and that required showing by the government only applies to those three provisions.

The rest of the bill does not require any showing of material harm to competition. So that's a big deal. But look, I do think there is some empirical evidence we can look to on self-preferencing and these sorts of restrictions if we look across the ocean to Europe and the DMA, who is running a very, very similar experiment.

And I think the reality is that the attempt to force Google, let's say, to de-integrate its maps, flights from its search engine have not had good effects for consumers. Consumers have a worse user experience in that it takes them longer to get the information and the answers they want. And also, frankly, a lot of traffic that used to go very directly to hotels, to restaurants, the things consumers are looking for, are now going to Google's intermediary competitors, right?

So, not helping businesses, not helping users, helping Google's competitors. Certainly, there are cases where self-preferencing might be anti-competitive. The antitrust laws in the United States can handle that. They do with Microsoft. But in general, this is common pro-competitive behavior that if you try and restrict it, I think it's going to have adverse effects.

One thing I would say, sort of on that line of ultimately helping competitors rather than helping consumers and innovation, that is really a concern when you talk about sector-specific legislation relative to the general framework that the Sherman Act provides, this idea of capture, right? If you're designing specific laws, more specific laws designed for particular industries, you get this problem of companies using that to benefit themselves, right? And especially in the tech markets where rents are very high, right, there's a lot of incentives for companies to use these sort of specific company, specific industry, specific legislations to benefit themselves.

So, you know, if the benefits of a sector specific regime, better rules, more tailored rules are not met, if this is actually going to be anti-competitive in a lot of its effects, and then you also have this new concern about regulatory capture, right? I just don't think that the sector specific logic here, justifies AICOA, right? There are certainly added costs of doing that. Just to be clear about something, I mean, I think Joel is right that we do need to step back and ask the question of, like, what is the market failure here? What are the problems that this bill is designed to solve, right?

I mean, whether it's an antitrust enforcement case or new regulation or legislation, where is the market failure? The way I would analyze that is saying, you know, are prices persistently high? Is output consistently declining? Is innovation dampening? And if you look at our tech sector, you just really don't see that in any sort of general way.

I mean, it's highly dynamic, highly growing. The big tech firms that would be captured by this bill invest billions in innovation every year. AI itself is evidence of the fact that creative destruction, dynamic competition are operative. So, you know, if we're looking for markets where there's market failure, I don't think that the tech industry really is a good example of that, frankly.

Which again asks, why do we even need this bill in the first place? Especially because, as Joel pointed out, the government's winning cases under the existing antitrust laws that we already have. And, you know, Joel raises the concern about cases taking too long. I could certainly agree with that, but you don't need something like AICOA to solve that problem, right?

There's a lot of other legislation you could think about to deal with that particular issue, rather than this new massive attempt to regulate technology companies.

Nancy Scola: Yeah. So, Joel, where is the market failure? Where do you draw the line between this is self-preferencing that's beneficial to users and this is self-preferencing that's egregious behavior? Is that something where you think, okay, you can, as legislators and regulators, you can step in and effectively detail that in this sort of sector specific, way that Joe talks about?

Joel Thayer: Sure. I think the answer is that intention matters, right? So, if you're self-preferencing because you really are just trying to give the consumer what they want, there's no problem there, right? And if you are using objective standards, which is also one of the affirmative defenses, right? If you say that you are actually applying, in AICOA, when you're applying your particular self-preferencing in a way that doesn't just simply favor you, there's a legitimate purpose behind it, then you're likely okay. Now, we can argue about the standard and whether the standard is too high or not, that's fine.

I think that will get negotiated as legislation moves along. But directly to your question as to, where is that line? I think the best way to answer that question is through a bit of an anecdote, and this happened actually with Google because Google really has done a pretty—look, they're first movers. They created search, and so they had the first mover's advantage, which already should give you a competitive advantage. But on the other end, you should be rewarded for the innovation that you have. Fair enough. But I think there was one instance in which innovation was directly impacted by Google's potential.

I would argue they've had monopoly power for a while. So, at this time, they certainly had it. So, as to innovation, which is also part of the consumer welfare series, which Joe and I both subscribe to, which is the general idea that, you know, bigness isn't necessarily bad unless you can demonstrate that there is some sort of market failure and market harm that impacts consumers adversely, which, totally on board with. I just think it's been too highly focused on price, not focused enough on quality of product and innovation. So, let's talk about the innovation side, because I think that directly answers your question. So, as to innovation, Google has a pretty weak case on it being the greatest innovator.

And especially when it comes to pushing back on the self-preferencing is really only there to give the consumer what they want. And as an example, Google used its extraordinary market power to kill and successfully slow the development of more efficient search technologies. And I'm speaking of course of vertical search, which was essentially getting created out in the UK. But they used their market power to essentially kill a company called Foundem. And Foundem gained users pretty fast, and Google recognized that. So, what Google did was, Google changed its search algorithm to digitally bury foundem.com by placing it as far down as 170 pages.

And irrespective of how many people directly searched for the product, mind you, and even if users actually typed in the name, foundem.com, you could not find it. Again, it's not clear, you obviously need cases to reveal this, and you need investigations to pull this out.

But what we saw from the Google search case, at least at the liability stage and through the trial, that this is precisely how it's using its self-preferencing. It's using its self-preferencing to actually potentially kill any other type of technology that would challenge its dominant position in search.

This is where I'll push back a bit on Joe's characterization of the government winning. I'm not sure a true win is just having a liability opinion. And even in the Microsoft case, we only had a liability opinion, somewhat similar to what we got in the Google search case, because as he knows and I think most people who are watching this knows, that the Microsoft decision was settled by a consent decree. It wasn't settled by a remedies decision. So, that's what makes it a little bit distinct from what we see in Google, where we have a remedies decision that basically says, "Hey, Google, we have you dead to rights. We found you as a monopolist." In fact, Judge Mehta minces no words. He literally says Google is a monopolist. That is a direct quote from the actual liability opinion. But then he says, "We can't do nothing about this."

Even though I agree with you that having these collusive, potentially collusive agreements with Apple is to make your search a default. We know you use that in an unlawful way. We know you use that to build a giant search empire. Yes, we know that you have been shafting smaller competitors from being able to access the same data sets. Yes, we understand that you basically use your market power to get into other individual verticals in order to create a giant pool of data up to you, so that way you control the entire internet ecosystem. Yes, we acknowledge that you're behind every website. But guess what? Can't do anything about it. I think that's where the call for Congress becomes more preeminent, where Congress has to be able to define what aspects of the market are an issue. And again, it's not just self-preferencing.

It doesn't say, "No self-preferencing for you big guys." And also, we talk about big guys, we're talking about companies that own a third of the entire market, which is already presumptively legal under current law. so that analysis is already consistent. And also, the fact that there is no market analysis, I push back on that a bit. It doesn't actually address the specific call in the statute that says that the self-preferencing or any of the other harms have to materially harm competition, and that too is defined.

When we're brought with a case, we can't just simply say, "You violated the statute." It has to be based at least in some factual groundings and some legal groundings. So, you have to do the investigation on the front end. You can't just simply say that you violated the law, and let's go to court and see what happens. Ask Lina Khan how well that went for her when she immediately just went to court without actually doing the necessary investigation beforehand. These statutes actually call for that investigation to happen before you bring any of those cases forward.

Because if you can't demonstrate that the self-preferencing actually materially harmed competition, you don't have a case on the front end. On the back end, you also have to make sure, even though it's an affirmative defense and that can be brought at trial, every smart lawyer knows, especially all the smart lawyers at the DOJ and at the FTC all know that those affirmative defenses are going to get raised. In fact, this law encourages the companies to raise those affirmative defenses. So, you are going to have to also do a market analysis for the court to head them off at the pass.

So, I think that the idea that you just bring these cases and then all of a sudden you get a win at the courts doesn't seem to comport with the text of the statute even as written. The affirmative defenses also require, well, I think a lot of economic rigor in order to determine that these things have actually done the things that the government says it does.

Then on the other end, I think the metrics that we're using in terms to define what a systematically important platform is, are pretty standard ways, we already use to determine market power. You still have to show market power based off of this subset of criteria that also seem to comport with how we already do it. And on the other end, you have to show that not only did they do the thing, but it materially harms competition on top of that. That requires another market analysis even before you get into court. So, to me, that particular framework makes it pretty unlikely that we're going to see the same issues that we saw in the EU.

And I think Joe and I have both lamented on how the EU has behaved in this market. I think we both written pretty extensively that the EU's DNA is probably not the way to go. And so, what I like and appreciate about this, and what I didn't appreciate about the first iteration of AICOA when it was first introduced, it kind of took that idea of that regulatory capture regime that Joe was describing, where it was up to the FTC to make the designation, and then you get to the actual harms.

This does the opposite. It actually follows a basic due process view. We're like, "Hey, we're telling you where you could go wrong in the market and what we're worried about. And also we're giving you the ability to defend yourself without having the agency de facto create this regulatory regime to label you as a gatekeeper until we decide otherwise."

We let the third-party arbiter figure that out or let us know if we're right or wrong on that, which is perfectly within the bounds of our checks and balances, which does not exist in the EU. So, I think that makes it very distinct, and I also think that pushes back on the idea that regulatory capture would be an issue here, especially when you have a third-party arbiter like a court deciding the designation on that particular firm, doing that particular thing, and also make a demonstration. Making sure that that the agency made a demonstration that this materially harmed competition.

Nancy Scola: Perfect. Joe, I wanted to ask you about the... I'm sure you see the results of the Google case a little bit differently. Can you explain, like, why you think that was not a sort of failure of the process.

Joseph V. Coniglio: Sure. But I do have to make a couple points first, if I may. So just on the facts, where the idea of counting revenues and counting users: that's exactly what the DMA does.That's the exactly the same thing that the DMA does to designate a gatekeeper, as what AICOA does. Now, that is also not the same, I'm sorry, Joel, as a market power analysis, right? AICOA defines a geographic market. Says the United States. But the other thing you need is a product market, and AICOA doesn't look at any product market. It just counts the number of users. And the issue with that, of course, is multi-homing, right? You could have somebody that is a user of Instagram and also a user of TikTok, right? We know that those compete, but they would still be counted as users, right? So just counting users is not a market share in a product market, and therefore it's not a market power analysis. It's not a dominance analysis like the Sherman Act, for example, would require. So AICOA follows the DMA exactly on that.

A couple other things. I'm glad to hear we both agree on the consumer welfare standard. What I would note is one of the hallmarks of the consumer welfare standard is allowing firms to show, especially in unilateral conduct cases or non-cartel cases, pro-competitive justifications for their behavior, right? Why something is going to benefit consumers. AICOA doesn't do that. AICOA doesn't provide you any general ability to show that consumers will benefit. It only provides for privacy and security. I think that is what the law says. Now, there are many other ways that conduct can benefit consumers other than privacy and security. And in fact, the courts have found that. Even in the Epic v. Apple case they found that under the Sherman Act, Apple's practices, not all of which would be prohibited by AICOA or subject to AICOA, have pro-competitive justifications for IP compensation, right?

Which would not be considered under AICOA. And on the material harm part, again, to be very clear, the government only has to show material harm for the self-preferencing obligations. For everything else, it doesn't have to show that. Instead, the initial burden is on the defendant to show that the conduct would not materially harm competition, to prove a negative. And that's going to be very, very difficult to do. And ultimately, to get a little technical, it's going to amount to what we call the modified per se rule. Whereas if a company has market power or it can't show it doesn't have market power, it's really not going to be able to defend itself, unless it can make out one of these two narrow pro-competitive justifications.

So, let me just be very clear about that. And Google was also not a first mover, by the way. I'm going through my notes. That was Yahoo, right? In fact, it's a common misconception. Many of these big tech firms were not the first movers in their markets, and it just goes to show you this idea of winner takes all really actually doesn't play out, in reality, frankly.

On the issue of again, consumers and innovation, all the upshot of that is these standards are way too strict and are going to end by not requiring the government to prove material harm, by having very, very high standards to show a very limited number of pro-competitive justifications. You're ultimately going to end up chilling a lot of pro-consumer and pro-innovation behavior that's going to make us worse off.

Nancy Scola: So, I do want to focus a little bit on this idea of the burdens on the companies, even though, these are big companies that are generally being targeted. We had an audience question of: how would the passage of AICOA impact leading American tech firms internationally?

So, the idea there might be you're placing a burden on these companies. It makes it harder for them to compete around the world. Other people argue, no, this is going to increase innovation in the US tech ecosystem. As a country, it makes the US better able to compete around the world. So, if you can address that question of, do you worry at all, Joel, that the idea of, okay, you are placing some new burdens on these companies, and it's just going to, like, slow them down when the US is trying to win this technological race? And then if you could bake in your response to Joe.

Joel Thayer: Sure. I would say that, even though I disagree with the DMA, the interest of the DMA has not slowed down these companies from getting bigger. And even in that marketplace, they're still getting large. I guess I credit that more to the failure of the EU to actually create a law that actually does the thing that they want it to do, as opposed to creating a framework that is actually workable.

But before I get into that, I want to at least address a couple of things that I heard Joe say. And one is that the AICOA does precisely what the DMA does. So, on the front end, the DMA, and as Joe knows, as we've all discussed, there is a front-end analysis that happens at the agency layer, right? Where the agency, or, well, I guess the European Commission basically designates these folks as gatekeepers, and they say, "We are basing it off of this criteria. That's how we're going to do it." Now, that's distinct from what AICOA does. AICOA basically has three factors. Not just, it's not just number of users, and not, it's not just how much money you make, but also it counts households, which is usually how you demonstrate market power in every other context, especially with respect to mergers.

That goes back to Philadelphia National Bank. That is straight case law, and that's 30%. What AICOA does is raise that to 34%, so already there's a difference. The idea that the affirmative defenses only allow for this, I'm just going to read one of the sections from affirmative defenses, just so that way we're all operating off the same facts. It doesn't just limit it to privacy. So, it actually says, under subsection B of affirmative defenses 2, it says, "No harm to competition. It shall be an affirmative defense to an action under paragraphs two through six," which is all the operative actions, "of subsection A if the defendant establishes by a preponderance of the evidence that the conduct has not materially harmed or would not materially harm competition."

Preponderance of the evidence is not the same thing as clear and convincing. So, all those things that the consumer welfare standard asks for, that Joe and I both think are good things to show pro-competitive effects, are actually baked into that and actually given a lower standard for the companies. Because preponderance of the evidence is, think of it more as a rational basis over, I guess if we're going to use the constitutional, strict scrutiny rhetoric. That is more in the rational basis territory than it is in strict scrutiny. So, already those affirmative defenses will allow a lot of these companies to basically say, "Hey, you understand that we're big. I understand that you made a case that we have this level of market power. But the government has the obligation to demonstrate, through clear and convincing evidence, that they've materially harmed competition." They have to get over not just that, but they have to get over an affirmative defense that says that is under a preponderance of the evidence standard, which is a much lower standard.

So, if anything, the companies do have the ability to avail themselves to all the same defenses that they would have. What I like about this, again, is that it gives structure to the courts. So, the courts have to actually evaluate whether or not these companies violated this portion of the statute and under these particular standards, and instead of just going off the rails and deciding that they want to be an economist for the day, or they want to be a technologist for the day. To me, I think that is a positive thing. I think that is an appropriate exercise of democratic power that we have in the United States, and I think that is an appropriate way to handle some of the issues that we all as conservative... Well, I speak as myself as a conservative. I don't want to speak for Joe.

What I like about this is that you have multiple checks here. You have, one, the legislator saying, "This is what we're worried about." Then you have the enforcement piece with the DOJ saying, "Okay, we have to build that case, and we're not going to leave it to you. We have to go to court to prove it." So, that is the third check. That is very different from what the DMA requires. The DMA is basically a top-down regulatory approach where it's just you're a gatekeeper, we are going to say that you did this, and yeah, you can challenge it in court, but the courts are very deferential in the EU, the European Commission, which is not the same as the United States, as we've all come to know.

And so those are all... those already, I think, cabin the issues with the harms to innovations that may or may not exist, but I would argue that the market is already broken. If you look at the way the market has been set up, like you have, like in every one of these individual sectors.

So, in the search market, it's pretty clear that there's one king that rules them all, and that's the court precedent. Then you have the app store market, which is run by two companies, basically Apple and Google, and you can make an argument that Microsoft exists in some universe, but I don't know anyone who has a phone with a Microsoft operating system that has anything. It's either Android or iOS. Those are the two options. When it comes to marketplaces, I mean, you really only have a couple. And even then, that’s not enough to make you liable for AICOA. You have to be able to demonstrate that you actually harm competition, which I, again, disagree with my buddy Joe that there is some element of... not even some element. There actually is, a rigor that you have to do in order to demonstrate a competitive analysis and an economic analysis that you, the government, are right, and you have to get that over a judge, and have to get that over an affirmative defense that has a very low bar of a preponderance of an evidence standard.

Nancy Scola: I wanted to give Joe a chance to respond. And Joe, can you include in your response this idea that it actually gives companies a little bit more certainty about what's allowable and what's not than they currently have?

Joseph V. Coniglio: So, I’m not sure about that. It's funny, one of the changes from the AICOA bill of yesteryear was that, under that bill, the DOJ and the FTC are required to put out guidelines, basically, at 270 days, I think, after it was passed, laying out, like, how the fines are going to be handled, some of these conduct restrictions. The implication of that being this is going to be really complicated, actually figuring out what all this means.

This bill doesn't do that, even though it's a lot of the same stuff. Instead, it just, somehow assumes courts are going to be able to figure this out in a year, which seems like it may be a recipe for summary justice, frankly. And if you look at the DMA, it's very, very similar, right? They have similar bans. But ultimately what happens is you have, in some cases, these very complicated specification proceedings, where you have to lay out exactly what this means, not to self-preference or data share or interoperability, right?

So, I don't think for a second this is going to be clear necessarily in any way. What I would say, I would just ask, every analysis of market power and monopoly power, you know, if you're doing it through indirect means requires demonstrating a product market, defining a product market.

AICOA does not require you to define a product market. So instead, it just counts users for your product, and that's not the same thing as any kind of product market or market power analysis. So I would just disagree fundamentally that AICOA has any dominance, market power or monopoly power analysis when you determine which products these rules are going to apply to.

Nancy Scola: Did you want Joel to respond to that or?

Joseph V. Coniglio: Sure.

Nancy Scola: Great. So, Joel, maybe briefly, what is Joe missing about the idea of how you're defining market power in the bill?

Joel Thayer: So, two things. I agree. If we want to clarify, have the DOJ create guidelines, I'm all for that. But just to be very clear, there is no statutory requirement for them to have merger guidelines, and yet they do that anyway, right? They want to be able to explain to the market how they're thinking about stuff. So. If the DOJ wants to do it, I guess there's nothing in the statute that prohibits them from doing it.

It also would seem that it would fall into common course with how the DOJ and FTC have basically handled a lot of these larger statutes that deal with markets. And so, to me, I'm like that's resolvable. Either we can add that in the statute, which I'm totally fine with, or I think most people would be fine with it, you know, requiring some sort of guidelines that come out.

But I don't think it's absolutely necessary because you don't actually need a statutory requirement to do that. They can just do it. So that's on the front end. The second was the question of this statute doesn't actually set a product market.

Now, the way I push back on that is that it goes back to the materially harms competition. “Competition in what?” is ultimately the question that the DOJ and the FTC are going to have to answer. And so that means that you're going to have to be able to describe, to some degree, where the anti-competitive behavior is, which means that you have to describe what market you're talking about. You can't just say, materially harms competition for the tech industry at large, the courts are going to be very skeptical of that. And I think you're starting to see that even in the statutory analysis requirements that we're getting from the courts.

So, the courts are asking every agency, "Prove your work. Show us your work. What does this mean?" So, when you say that they're harming competition, harming competition in what? It would be very unwise for any DOJ to look—or FTC—and say, "Google violated this statute because they materially harm competition," without deciding which markets you're talking about, what, facts have to go in or out. Joe and I have been probably on the receiving end of getting yelled at and chastised by a judge, and I can tell you, you don't want to be in that position if you have not answered those questions going into it. Now, if what Joe is saying requires more clarity, I think that, again, this is a discussion draft.

If that's something that you want to raise with the legislator and the legislator feels that's valuable, that's something even I would support, great. Let's have that conversation. Materially harms competition already gets you 90% to even 95% of the way there.

Nancy Scola: So, Joe, can I kick it back to Joe? Why aren't you comforted by that those definitions would get worked out in the legal process?

Joseph V. Coniglio: Sure. So, to be clear as Joel, I think, conceded, AICOA does not require you to define a product market in a way that you typically would under the Sherman Act to show market power, monopoly power, or dominance.

And it's going to apply those restrictions notwithstanding, right? With respect to the competitive effects analysis, which is separate from the market power analysis about defining a monopoly or whatnot, Joel points out correctly that there is this language, as I've conceded, that says that for those instances where the government doesn't have to show material harm to competition, I would note for self-referencing, the bill doesn't allow you to say it's not going to materially harm competition.

The government gets the last word. If it proves material harm, all the company can do is make the privacy/security justifications. It actually can't come back and say, "No, it's not going to materially harm competition." But for those other instances where the initial burden is on the companies to prove that it's not going to materially harm competition, that is going to be extremely difficult to do because typically the way that works under the Sherman Act is the government comes up with a theory about harm to competition, and then the company shows why that theory is wrong and it's not going to materially harm competition.

But asking a company to prove the negative, that it's not going to materially harm competition, in principle you'd have to find a way to rebut every possible theory of harm that the government could come up with, right? Now, in reality, the only way a company's going to be able to do that is if it can show that it doesn't have market power, right?

If it can prove that it doesn't have market power, therefore it doesn't have the ability to materially harm competition, right? And this is different from showing a pro-competitive justification, right? If a company says, "Well, we're not going to materially harm competition because there's a pro-competitive benefit here," that's not sufficient, right? Because there could be anti-competitive harms that would await or would outweigh that pro-competitive benefit.

So, that's the distinction between a denial and a defense in the law. So, basically if a company can show that it does not have market power and it can't make out one of these defenses, its conduct is going to be illegal. And that rule in antitrust is basically known as the modified per se rule that we used to apply to tying, where if a company has market power, the tying is basically illegal. And that sort of a rule applied to practices like refusals to deal like, using your data to come up with a new product, right?

That's going to chill a lot of pro-competitive behavior, because in most cases, even if a company does have market power, there can be pro-competitive benefits that result from that, right? And if this really was a consumer welfare standard where companies could present real pro-competitive justifications for their behavior, it wouldn't just be limited to privacy and security.

You would allow firms to make all the other types of pro-competitive justifications they have, which is totally different, in my opinion, from proving that it won't materially harm competition. That's a different analysis, and, again, in both cases very, very different from how the Sherman Act would treat the legal analysis, let's say, under the Sherman Act.

Nancy Scola: Joel, do you have a one-minute response and then I want to start wrapping things up?

Joel Thayer: Sure. I just want to make a factual distinction, and again, you don't have to take my word for it, you don't have to take Joe's word for it. I encourage everyone to actually go read the bill. On the idea that the self-preferencing doesn't have that limitation on materially harms competition, I'm looking at the statute right now in all three instances, which is subsection 3A1 through A through C, all has the phrase "would have to materially harm competition," which again is a high standard.

So, on that end, I disagree with that. Joe keeps referring back to, like, privacy's the only thing that matters, or there's only this one little thing. I've already read to you that there is at least another one, and there's actually three affirmative defenses and they all imbue this idea of showing pro-competitive effects in this market. Now, on the issue of whether or not we absolutely need a product market, I think that the statute just basic statutory interpretation, this is high-level stuff, like almost one-off stuff.

If you are going to say that the statute's trigger is that it has to materially harm competition, you have to ask competition in what? You can't just simply say that you're harming competition across the board. And so again, I think that's something that's been lost a bit in the conversations when coming to AICOA.

Again, there's no product market described in Section 2 of the Sherman Act. There's no product market described in any of the antitrust laws. They're general application of statutes, so why is it okay not to define a product market in these general statutes, but when we actually do a specific statute that's trying to address an industry that has a lot of vertical integration in it and has a lot of products in it, why can't we also use those same types of categories, or that same type of language, in order to create the same type of analysis, just in a tailored way to the tech markets. It just seems like antithetical to traditional statutory interpretation, and just seems like we just want to go back to doing this wide-ranging general statute regime that hasn't really yielded any of the benefits that consumers are aching for, which is actually more competition and more diversity in services.

Nancy Scola: Okay. Joe, very quickly, like a 20-second response there.

Joseph V. Coniglio: Of course. So, on self-preferencing and this area of no harm to competition, which as Joel notes is affirmative defenses B, section 2, it specifically states, "It shall be an affirmative defense to an action under paragraphs two through six," right?

Under two through six. That specifically excludes paragraph one, which is the self-preferencing part. So, you don't have the ability to show no material harm to competition for self-preferencing. That's what the statute says. Only the government has to show harm to competition, but then the company can't deny it, right?

Which is not how the Sherman Act works. With respect to product markets under the Sherman Act, when you restrict unilateral behavior, which is what most of this is, you have to prove that the company has a monopoly position or a dominant position with an intent to exclude. You can do that by defining a product market and showing a high market share and barriers to entry, or you can do that by showing direct evidence of monopoly power or dominance.

AICOA does neither. Just showing that a practice may have material harm in certain markets is not the same as showing direct evidence of monopoly power in that market or direct evidence of dominance.

Nancy Scola: In the generosity of spirit to engage in a little thought experiment, what is the best argument against your position in one or two sentences? So, Joel, what is the best argument against why we need AICOA? And then Joe, the same question to you.

Joel Thayer: Yeah. Every lawyer likes to argue against themself. I think the best argument would be that, you know, a lot of these antitrust cases whether it comes to Google or Apple, we're still in the appellate stage, like in the Google Search case.

We still have to see exactly where the DC Circuit will land in these cases. There's all these other appeals that are going to happen. It might be true that we live in a universe where Joe's right. Where the antitrust laws as they are currently written work.

Nancy Scola: Yeah.

Joel Thayer: I will add just one clarification. So again, my argument on the self-preferencing part is that the materially harms competition, there is a defense that's imbued in it, and I totally credit the argument and I can see the argument that the statute does not allow for those affirmative defenses.

I thought we were most more or less talking about the ties and all the other aspects of it, but we're talking about specifically the self-preferencing. I'm saying that you can bring those defenses when you are arguing against, like, that's the whole point of a court proceeding, an adversarial proceeding.

You're allowed to bring counterfactuals to the court. But I think the strongest argument against my position would be that, hey, the courts are still working through this. There's still appeals that have to happen, and maybe we should slow our roll. That would be, I think, the best steelman against mine.

Nancy Scola: Okay, thank you. Joe, same question.

Joseph V. Coniglio: So, the strongest argument in favor of Joel's position— I want to thank him for doing this debate, I really enjoyed this discussion— is that there is sort of a loophole in the Sherman Act around this idea of monopoly leveraging that I talked about, where if you have a monopoly in one market and you use that to get an advantage in another market, like materially harm competition in another market, right?

But you don't get a monopoly, then you can't really bring that case under the Sherman Act, right? And that's, you know, in the European Union, you can have an abuse of dominance where you do have that leveraging. You have a dominant position in market X, you use it unilaterally to benefit yourself in market Y, that can be an offense. But the thing is, we don't need AICOA to solve that problem, right? In theory, section 5 of the FTC Act could do that sort of thing if you want it to, and frankly, I don't think we should have a monopoly leveraging theory because just because you're getting an advantage in another market doesn't mean there's going to be harm to consumers, right?

Doesn't mean there's going to be any real economic harm we should be concerned about. Take Microsoft, let's say Microsoft using its ecosystem to prop up its browser, right? Edge. Yeah, it may get an advantage in Edge, but it's competing, right? There's no necessarily harm to competition. It could actually be enhancing competition by helping it better compete with Safari and Apple. So I agree there is a concern about monopoly leveraging not being really something we can deal with, but I don't think this is the way to deal with that problem.

Nancy Scola: Okay. Just to wrap up, we've about 30 seconds left, so very briefly, is this even something to worry about? Is this bill going anywhere? Joel? Not what you hope to have happen. What's actually going to happen?

Joel Thayer: Sure. So, look. First, I want to thank Joe for giving me this opportunity. It's big of him, it's big of ITIF to invite folks with contrary views onto these platforms.

I think this is the way we get healthy discourse, and it's certainly the way we get good policy. And so on that front of good policy, look, clock's a-ticking. We don't have much time on the congressional calendar. And my instinct is that, there's a lot of things that have to get moved over. You have child safety in the tech space that seems to be taking a little more precedence than maybe the antitrust stuff. But again, anything could happen. There are must-passes where I thought, you know, there was no way a bill was going to get passed, and this gets thrown in there, so maybe that might be the answer.

But, I am cautiously hopeful.

Nancy Scola: Okay. Joe, same question.

Joseph V. Coniglio: I don't think so. And I think one of the telling things is that I think the number of Republican co-sponsors on this bill has decreased every time that they've brought it, which suggests, I think rightly, that certainly Republican voters, this is not an area where they're focused on.

They're focused on growing the economy, affordability, making sure America remains the leader in the world. They're not worried about Google integrating their search and their map service. In fact, again, one of the reasons why these services are so widely used, and so key to our economy, is consumers like them.

So, I just don't think this is something that's going anywhere.

Nancy Scola: Okay. Excellent. It's sort of cliched at the end of these panels to say we could go another hour, but I feel like you two could easily go another hour or two, so. But, I want to thank you, Joe and Joel. At the risk of editorializing, I commend you on being willing to have a vigorous debate with people you disagree with.

So, thank you both for that, and thank you for ITIF for hosting this, and thank you for all of you watching. Great.

Joel Thayer: Thank you.

Joseph V. Coniglio: Thank you, guys.